Abstract
Little progress has been made towards an understanding of the aetiology of postpartum psychosis, despite the longstanding assumption that it is precipitated by biological factors. Methodological problems in this area of research may be responsible for this failure, as outlined here using the example of the relevant research into the role of oestrogens. A solution to these problems is discussed, i.e. the use of a homogeneous high-risk subject sample, which would satisfy the requirements of future research and offer several advantages over previous work.